Latest Notable AI accomplishments

  • Topic: AI 
  • Thread starter gleem
  • Start date
  • Tags Tags
    Ai
155 replies · 38K views · last reply
Matterwave said:
https://github.com/yang-mills-group/YM/blob/main/docs/assessment-1.pdf

It looks like they "validated" the results. With a bunch of caveats. I haven't gone over / understood even the validation LOL.
Peter Woit (https://www.math.columbia.edu/~woit/wordpress/?p=15986) is hinting at the fact that they proved weak-existence (compactness) and that's why they are not claiming to have solved the whole thing yet. Also he wondered earlier if they just did SU(5) GUT, without any Higgs sector.
 
Last edited:
misterknister said:
I don't understand the "math and physics are dead" outcry. Even if AI becomes an all-knowing science oracle, people will still need to know math or physics to make sense of the answers, don't they? Or to ask the right questions.

It's the same with coding. Coding isn't dead either and that's being supposed since the dawn of AI.
My worry with this is that coding is still necessary for industry while academic research is sometimes far from it, specially pure mathematics. Academic research is very badly funded, imagine what will happen now that they know a single researcher can do the work of ##N## researchers with the help of AI. Even in coding the salaries and positions are getting reduced.
 
misterknister said:
Let's say in the future AI supposes a solution to Quantum Gravity. It gets some predictions right, but nobody understands the equations. So everybody just says: "Well, cool. So this is it. Let's have lunch".

I don't believe that.
Neither do I. There are thousands of academic scientists. They will not accept anything on faith. Even their most respected colleagues' explanations are scrutinized to the fullest, and they will do the same to AI to make sense of whatever it produces.

The rest of the population is another issue. They either lack the interest to care or do not have the ability or training to evaluate AI's revelations.

The challenge may be trying to recruit enough talent into the human led scientific community.
 
gleem said:
The challenge may be trying to recruit enough talent into the human led scientific community.
Getting talented candidates for recruitment might not be an issue but getting the funding will.
 
pines-demon said:
Peter Woit (https://www.math.columbia.edu/~woit/wordpress/?p=15986) is hinting at the fact that they proved weak-existence (compactness) and that's why they are not claiming to have solved the whole thing yet. Also he wondered earlier if they just did SU(5) GUT, without any Higgs sector.
I can't comment on the physics as Woit and everyone on Gross's committee obviously know their stuff way better than me. But Jaffe himself is on the committee. So the point about Witten-Jaffe's write up excluding compactness-based proofs seems a bit.. odd? The guy who wrote it is on the committee and backs the proof. 😂
 
Matterwave said:
I can't comment on the physics as Woit and everyone on Gross's committee obviously know their stuff way better than me. But Jaffe himself is on the committee. So the point about Witten-Jaffe's write up excluding compactness-based proofs seems a bit.. odd? The guy who wrote it is on the committee and backs the proof. 😂
The collaboration is NOT saying that they solved the Clay Millenium Problem. They are claiming that they solved part of it. Woit is commenting that maybe they solved the part that Jaffe said that does not count to claim the prize.